A system can determine a first coordinate location and a second coordinate location collectively representing a spatial positioning of a graphical object in a user interface. The system can determine a first pitch for a foreground sound based on the first coordinate location. The foreground sound may be a synthesized voice describing textual content associated with the graphical object. The system can determine a second pitch for a background sound based on the second coordinate location. The system can generate an audio signal configured to cause an audio device to output a composite sound that includes the foreground sound at the first pitch and the background sound at the second pitch.